image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

image.png

We build a model with augmentation and without dropout:

We build a model with both dropout and augmentation:

We will try to get a model with VGG16 & using the imagenet with the data augmentation and the dropout layer

This is the best model we were able to get to!

The Test accuracy is 0.9800000190734863 and the Test loss: 0.21156862378120422 !